[phenixbb] Geometry Restraints - Anisotropic truncation

Dale Tronrud det102 at uoxray.uoregon.edu
Thu May 3 10:43:44 PDT 2012

   Yes, "never" is too big a word for this, but writing "nearly never" simply
begs for explanation.  The rejection test that Pavel and Randy perform is
extremely conservative and individualistic.  In addition, the last time I
looked in a Phaser log file the excluded reflections were individually listed
and, as you said, the exile is not permanent - there is a parole hearing
every round of the calculation.

   This is very different than the mass rejections being discussed here.


On 05/03/12 10:23, Nathaniel Echols wrote:
> On Thu, May 3, 2012 at 10:05 AM, Dale Tronrud <det102 at uoxray.uoregon.edu> wrote:
>>   An important point is that the Fc's must never be used to judge the
>> quality of the Fo's in a production environment.
> I'm not sure that's completely fair - they can certainly be used to
> identify Fobs values that are wildly at odds with expectations.  This
> is what Pavel does with outlier rejection in phenix.refine (I forget
> the exact reference for this protocol but I think it's Read 1999 or
> something like that).  The difference, I think, is one of degree, and
> also that (at least in the current version) the outliers are
> recalculated at each cycle of refinement and never permanently
> excluded, unlike anisotropic truncation.
> -Nat
> _______________________________________________
> phenixbb mailing list
> phenixbb at phenix-online.org
> http://phenix-online.org/mailman/listinfo/phenixbb

More information about the phenixbb mailing list